The objective of the Biostatistics Core is to help make quantitative inferences based on research about the causes and control of diabetes and related endocrine and metabolic disorders by consulting and collaborating on relevant research. The Core provides experimental design, data management and data analysis in support of approved projects in the Center. Core personnel also develop novel biostatistical techniques, including models in statistical genetics, that are necessary to improve the analysis of such data. Examples of such biostatistical techniques include; improved estimation of hormone levels in sera from assays, including RIA's; development of a model to identify pulses in hormone secretion from a sequential series of assays; improved methods to include data from dropouts in 'intent to treat' analyses of controlled clinical trials; methods to combine information from multiple endpoints into a single overall test of efficacy; and analysis of pedigree data with emphasis on linkage analysis. Members of this Core will provide expertise in developing additional models in collaboration with investigators from the Michigan Diabetes Center.